Caller ID Strangeness

[TromboneAl]

I just got a call from someone whose caller ID told him that I had called him a minute earlier. However, I hadn't made any phone calls. He said he thought someone was using my number or something.

I figure the Caller ID was just wrong.

Anybody know anything about that?

 

[Sandy]

As long as they didn't ask for your SSN, credit card info, and such, I wouldn't spend any time on it. I would assume he read/wrote the number on his caller ID incorrectly. Or maybe dialed incorrectly.

 

[eridanus]

There are some internet sites that let net callers fake their number. So even if caller ID says the call is from 1-800-CITIBAN, it may not be.

https://www.itellas.com

 

[RunningBum]

Telemarketers are starting to use this, to keep from getting detected and penalized for violating the no-call lists. The one from "card holder services" or something like that saying you can get a better rate is one of them.

 

[sailor]

Caller ID is easily "spoofed"

Or shall we rather say, easily set by software - and there are good reasons.
A lot of call centers have their caller IDs "dynamic".
For just a single call I would not worry about it, but if there is a good reason you can request a copy of a billing record from your carrier, to confirm it was not your phone.

It would be a very rare occurence that actually transmitted digit would be read wrong by the end user phone, but theoretically it could happen.


sailor,
who plays with this stuff for living and has 55 T-1s terminating on his desk (about 1200 phone lines)

 

[ERD50]

http://www.edn.com/index.asp?layout=articlePrint&articleID=CA56628

this is a link to EDN - it keeps coming up 'advertisement'


that link says an 8-bit checksum is included. I don't know how common errors are in transmission, but I think an 8-bit checksum would mean that there is a 1 in 256 chance that the error could also randomly match the checksum.

IOW, it could happen. I also wouldn't worry unless it was repeated, or if they were suspicious calls.

-ERD50

 

[sailor]

http://www.edn.com/index.asp?layout=articlePrint&articleID=CA56628
I don't know how common errors are in transmission

Order of magnitude Bell 202 over POTS BER in US is about 1e-06, so one bit error for one million bits transmitted.
Typically better in metro areas with newer equipment and worse in rural areas further you get from CO.

 

[Telly]

sailor,
who plays with this stuff for living and has 55 T-1s terminating on his desk (about 1200 phone lines)

Now are those actually T-1's, or are they really just DS-1's? Many people use the names interchangeably, which is wrong. Nothing like a long repeatered span to teach people the difference, as they get up from getting knocked on their *ss!


Back to the CID situation, doesn't internet telephony allow you to set your own outgoing number appearance? I don't have internet phone, but heard that somewhere.

 

[ERD50]

So, 1 in 1,000,000 bit errors.

10 digit phone numbers take 30 bits to represent, so 1 in 33,333 phone numbers transmitted would have an error, Most of those would just display 'ERROR', there is error detection, but no error correction in CID.

And, 1 in 256 of those would not be detected. Undetected errors, 'wrong numbers' would be a one in 8,533,333.33 occurrence.

Assuming I can do math on 0.5 cups of coffee .

-ERD50

 

[sailor]

Now are those actually T-1's, or are they really just DS-1's? Many people use the names interchangeably, which is wrong. Nothing like a long repeatered span to teach people the difference, as they get up from getting knocked on their *ss!

A mixture of both and more (6 real Ts, 42 DS1s (or if you insist on using "proper" names I should say DSX-1s) and 7 120ohm E1s.

Back to the CID situation, doesn't internet telephony allow you to set your own outgoing number appearance? I don't have internet phone, but heard that somewhere.

Some do, some don't. Depends on carrier.

 

[sailor]

So, 1 in 1,000,000 bit errors.

10 digit phone numbers take 30 bits to represent, so 1 in 33,333 phone numbers transmitted would have an error, Most of those would just display 'ERROR', there is error detection, but no error correction in CID.

And, 1 in 256 of those would not be detected. Undetected errors, 'wrong numbers' would be a one in 8,533,333.33 occurrence.

Assuming I can do math on 0.5 cups of coffee .

-ERD50

Math is correct, but assumptions flawed
Both formats used in US to transmit caller ID data use ASCII encoding for phone numbers. So one byte per digit. Also IIRC 8N1 is used, so to transmit one byte 10 bits are used.
So 100 bits for 10 digits phone number.